Missing collection strata permit both signs of flower pigment change in six herbarium samples from 1941 to 2017

We tested whether early and late flower collections support change estimates within the same species, month and location. We used 1,235 specimens from a public archive. Six separate samples cover Australia, Eurasia and North America in 1941-1985 and 1986-2017. Only 11-46% of pooled collection weight had records in both halves within species, month and 5-degree cells. With observed cell means held fixed and absent means allowed from 0 to 100% pigment area, all six sharp bounds include negative and positive changes. Finer grids reduce overlap. These are archive support bounds, not confidence intervals or causal climate effects. They show the extrapolation needed for a fixed collection target. They do not establish population change or its absence. One archive, broad cells, sparse records and unknown collection effort limit use. The package gives code, source hashes, aggregate results, numerical checks and a figure. Total-column ozone is not surface air pollution; no pollution effect or weather extreme is tested.

herbarium · floral pigmentation · observation bias · common support · partial identification · climate change